One aim of upcoming high-intensity laser facilities [1][2][3] is to provide new high-flux gammaray sources [4]. Electromagnetic cascades [5-9] may serve for this, but are known to limit both field strengths and particle energies [10], restricting efficient production of photons to sub-GeV energies [11][12][13]. Here we show how to create a directed GeV photon source, enabled by a controlled interplay between the cascade and anomalous radiative trapping [14]. Using advanced 3D QED particle-in-cell (PIC) simulations [15] and analytic estimates, we show that the concept is feasible for planned [3] peak powers of 10 PW level. A higher peak power of 40 PW can provide 10 9 photons with GeV energies in a well-collimated 3 fs beam, achieving peak brilliance 9 × 10 24 ph s −1 mrad −2 mm −2 /0.1%BW. Such a source would be a powerful tool for studying fundamental electromagnetic [16] and nuclear processes [1, 17,18].Advances in high-intensity laser science offers opportunities for creating a new kind of high flux gamma-ray source, based on the use of strong laser fields to accelerate particles and stimulate emission within a single optical cycle [11-13, 19, 20]. However, from a naive consideration of particle dynamics one would expect particles to be expelled from the temporal and spatial regions which are optimal for energy gain (i.e. the electric field antinodes). Furthermore, for intensities above 10 24 W/cm 2 radiation losses prevent particles from reaching their potential maximum energy (during a single phase of acceleration). This limits the effective generation of photons to sub-GeV energies. [10,21] Our aim here is to find the optimal strategy for source creation. To do so we exploit the anomalous radiative trapping [14] (ART) of electrons and positrons in a dipole wave [22], the latter being the field configuration which provides the highest possible field strength for a given peak power arXiv:1610.06404v1 [physics.plasm-ph]

We present a comprehensive and systematic investigation of the fundamental physical limitations of Faraday isolation performance at high average powers that are due to thermally induced birefringence. First, the operation of various Faraday isolator designs by use of arbitrary orientation of cubic magneto-optic crystals is studied theoretically. It is shown that, for different Faraday isolator designs, different crystal orientations can optimize the isolation ratio. Second, thermo-optic and photoelastic constants for terbium gallium garnet crystals grown by different manufacturers were measured. Measurements of self-induced depolarization are made for various orientations of crystallographic axes. The measurements are in good agreement with our theoretical predictions. Based on our results, it is possible to select a crystal orientation that optimizes isolation performance at high average powers, resulting in a 5-dB enhancement over nonoptimized orientations.

Two methods of compensation of thermal lensing in high-power terbium gallium garnet (TGG) Faraday isolators have been investigated in detail: compensation by means of an ordinary negative lens and compensation using FK51 Schott glass possessing a negative. Key thermooptic constants for TGG crystals and FK51 glass were measured. We find that the contribution of the photoelastic effect to the total thermal lens cannot be neglected for either TGG or FK51. We define a figure of merit for compensating glass and show that for FK51, an ordinary negative lens with an optimal focus is more efficient, but requires physical repositioning of the lens for different laser powers. In contrast, the use of FK51 as a compensating element is passive and works at any laser power, but is less effective than simple telescopic compensation. The efficiency of adaptive compensation can be considerably enhanced by using a compensating glass with figure of merit more than 50, a crystal with natural birefringence or gel.

The use of two-dimensional Bragg resonators of planar geometry, realizing two-dimensional (2D) distributed feedback, is considered as a method of producing spatially coherent radiation from a large sheet electron beam. The spectrum of eigenmodes is found for a 2D Bragg resonator when the sides of the resonator are open and also when they are closed. The higher selectivity of the open resonator in comparison with the closed one is shown. A time-domain analysis of the excitation of an open 2D Bragg resonator by a sheet electron beam demonstrates that a single-mode steady-state oscillation regime may be obtained for a sheet electron beam of width 100-1000 wavelengths. Nevertheless, for a free-electron maser (FEM) with a closed 2D Bragg resonator, a steady-state regime can also be realized if the beam width does not exceed 50-100 wavelengths. The parameters for a FEM with a 2D planar Bragg resonator driven by a sheet electron beam based on the U-2 accelerator (INP RAS, Novosibirsk) are estimated and the project is described.

Theoretical investigation of a short electron beam (extended bunch) interaction with a backward wave propagating in a slow wave structure demonstrates the possibility of producing ultrashort superradiance pulses with a peak power which exceeds the power of the driving beam (conversion factor K>1). It is shown that a nonuniform slow wave structure with optimized profile is beneficial in order to increase the conversion factor. The results of theoretical analysis are confirmed by the experiments. At X band using the SINUS-150 accelerator (4 ns, 330 kV, 2.6 kA) 0.6-0.8 ns superradiance pulses with a peak power of 1.2 GW and a conversion factor of 1.5 were obtained. Similar experiments at Ka-band based on the RADAN-303 accelerator (1 ns, 290 kV, 2.5 kA) demonstrated production of the superradiance (SR) pulse with duration 200 ps and peak power about 1 GW (conversion factor of 1.4).